The early stages of osteoarthritis (OA) are associated with loss of proteoglycans (PGs), breakdown of the collagen matrix, and change in water content. Magnetic resonance imaging (MRI) is routinely used in the diagnosis of OA because of its high spatial resolution and excellent soft tissue contrast. Recent research has focused on establishing correlations between quantitative MRI measurements (T1, T2, T1r, and water content) and the biochemical properties of articular cartilage. While emphasis has been placed on the changes seen in cartilage, OA is a multifactorial disease involving different tissues and when one joint tissue deteriorates, it is likely to affect others and contribute to failure of the oint as a whole. A particular problem arises in MRI assessment because many joint tissues such as menisci, ligaments and tendons have T2s of only a few milliseconds. As a result they show little or no signal with conventional clinical spin echo (SE) or gradient echo (GE) sequences, which have typical echo times (TEs) of several milliseconds or longer. The lack of signal means that it is difficult or impossible to accurately measure their T1, T2, T1r and water content. Furthermore, water is present in both bound and free compartments within joint tissues. The bound water components have shorter T2s and are usually inaccessible with conventional clinical pulse sequences. We have developed UTE sequences with nominal TEs of 8 s which are 100-1000 times shorter than the TEs of conventional sequences, and allow us to detect water signals from MR invisible tissues in the knee joint. In addition we have developed a spin-lock prepared UTE sequence to measure T1r, a T2-prepared UTE sequence to measure T2, a UTE PD sequence to measure water content, and UTE bi-component analysis to quantify the fractions of bound and free water components in the principal tissues of the knee joint. In this proposal, we will evaluate the sensitivity of both UTE and clinical sequences for evaluating PG depletion as well as changes in collagen microstructure and water content by studying relatively normal cadaveric patellae (n=40) and menisci (n=40) before and after sequential enzymatic treatment (Aim 1). Then we will compare UTE and clinical sequences for quantitative diagnosis of OA in cadaveric knees with normal (n=20) appearance as well as mild (n=20) and moderate (n=20) disease (Aim 2). Finally we will characterize patterns of knee joint degeneration in a cross sectional assessment of four groups of human subjects: normal controls (n=20), patients at risk of OA with knee pain but normal radiographs (n=20), patients with mild OA (n=20), and patients with moderate OA (n=20). We will correlate the UTE and conventional MR measurements with Kellgren-Lawrence, WOMAC, Tegner-Lysholm and IKDC clinical scores (Aim 3). Successful completion of the proposed work will provide new opportunities to characterize OA in a much more comprehensive and systematic way than has been possible with conventional clinical pulse sequences. This is likely to have a major impact on early detection in OA, monitoring disease progression, and assessing response to therapy.